Libor) -- the average rate of interest at which banks lend money to one another (on which see, eg Huff Post's "Barclays Libor Scandal: Rate Rigging Affects Your Loan Payments" here). The Libor scandal has generated another round of bank-bashing in general and has seen the sudden departure of some senior figures from their positions of remarkably profitable employment. With all this going on, someone in the bank's marketing department must be cringing at its poster-ad campaign (right) -- an unfortunate exercise in "truth in advertising", one might say [Thanks to those readers who have written in to let the Kats know that the advertisement is a spoof -- and congratulations on making the effort to check!].
here, but he is the first to concede that this was never going to be the last word on the subject. Yesterday his friends at MARQUES, the organisation for trade marks in Europe, set out its own carefully thought-out position on the problems of using class headings as a way of specifying goods and services for which registration is sought. You can read MARQUES's position on Class 46 here. It has been a big week for blog debutantes: Eleonora Rosati and Kate Manning opened their blogging accounts here on the IPKat, while Iona Harding produced two promising posts for the copyright-oriented 1709 Blog which you can check out here and here. Meanwhile, on the Class 99 design law blog, you can find an introduction to Open Product, a curious open licence scheme for designs that are not intended to be exploited commercially, here.
IPKat team member Neil's post last month under the title "IP Practitioners: Ignore Ethics and Moral Philosophy at Your Professional Peril" (here) generated numerous responses; here's one from Colin R Davies (Senior Lecturer, Intellectual Property Law, University of Glamorgan):
While AI experts are trying to
make bots think like humans,
we already have a Bot who
does just this -- the CJEU's
Advocate General Bot
"Neil’s blog on the issue of human responsibility for the actions of auto bots making decisions independently of the human operator raises a further question, fundamental to the field of intellectual property, in relation to such “creatures”. As we move deeper into the 21st century (boldly going where no man has gone before), having just passed the 100th anniversary of the birth of Alan Turing, commonly regarded as the founder of modern computing, the ability of a computer to “think” and develop ideas autonomously and independently, without the need for any human intervention or input, is becoming a reality. There is emerging a new breed on this planet – fully developed and functioning Artificial Intelligence systems capable of obtaining, assimilating and utilising data without any human control.
Turing postulated that the human brain was simply a machine, one which contained a series of “gates” that could be turned on and off to make decisions. He firmly believed therefore that this could be replicated by a machine and a truly artificial intelligent computer system, equal in computing power to the human brain, could be developed. Such a system would in essence “think”.
Although the jury is out as to whether we will succeed in building a truly artificial brain, the fact is that artificial intelligence systems exist controlling many facets of our lives... and they are becoming more intelligent.
So who, if anybody, has a claim to any intellectual property created by such AI systems? This is not fiction – such systems have written books, music and even designed patentable inventions. The issue is not new and was addressed in the UK by the Whitford Committee back in 1977. However, not surprisingly, in those days the committee saw these (then primitive) computers as nothing more than tools and consequently recommended that IP rights in any computer generated works should belong to the person by whom the arrangements necessary for the creation of the work are undertaken. This was embodied in Section 9(3) of the Copyright, Designs and Patents 1988. Interestingly, no such provision found its way into UK patent law. Nor anywhere in any of the UK legislation is there any specific requirement for the creator/holder of IP rights to be human.
However, things have moved on apace in the field of artificial intelligence since the days of Whitford. Present-day systems are far removed from those at the time of Whitford and are making decisions in their own right and rewriting their original programming as they learn from their mistakes further developing their own intelligence. So should we cling on to such outdated concepts? The question needs to be addressed as to ownership of any IP rights created by such systems.
There are several human stakeholders who might lay claim to any rights created by such AI systems – the creator of the original program, data operator, owner/investor, operator. All have potential claims, albeit nebulous, but in my view the arguments against granting them such rights outweigh those in favour. A full analysis of their respective claims lies outside the scope of a blog but suffice to say that the fundamental problem in granting any of them such rights is that they would not have had any input or involvement in the final work (for a more detailed analysis see the author’s article referenced below).
So what are we to do? In the absence of any tangible claim by a human being, are we to leave such rights to fall as orphan works? Or simply deny any protection? Both cut across basic reasoning as to why we have intellectual property rights. Or should we perhaps consider a more revolutionary solution, not exactly the creation of a robotic Bill of Rights, but the creation of a new legal persona to initially hold such rights. The AI system itself? Bizarre? Perhaps, but no more so than the granting of legal persona, as has already been done, to a fiction – the limited company. These questions need to be answered.
What do people think? Are there any AI computers out there with an input? For the writers views see "An Evolutionary step in Intellectual Property Rights, Artificial Intelligence and Intellectual Property", Computer Law & Security Review 27 (2011) 601-619".